The treatment of brine concentrates from reverse osmosis (RO) processes used for water recycling applications poses a major challenge. This research addresses the use of several bio-physicochemical processes for removing ammonia, nitrate and sulfate from RO brine rejects, and essentially consists of four different stages. The first stage involves nitrification of ammonia using a bioactive fluidized-bed reactor (FBR) process under aerobic conditions. The second stage deals with denitrification of effluent from the first FBR in a second FBR operated under anaerobic conditions. The third stage achieves simultaneous reduction of sulfate in the FBR performing denitrification. The final stage addresses the gas-phase biofiltration of hydrogen sulfide formed in the third stage. Among various phases of the treatment process, this work focuses the modeling and design of the denitrification and simultaneous sulfate reduction stages of the treatment with emphasis on process upscaling and design.

A mathematical model was developed for predicting the dynamic behavior of the FBR denitrification process in two stages along with simultaneous sulfate reduction under anaerobic conditions. The model involved the application of sorption and biodegradation phenomenon in an FBR system addressing liquid film and biofilm transport besides diffusion into the sorbent granular activated carbon (GAC) media particles. The modeling approach was also applied to non-adsorbing bioactive media particles such as sand. The model was intended to simulate or predict the dynamics and performances of FBR systems for denitrification and sulfate reduction under a variety of process and operating conditions. Sensitivity studies were undertaken to evaluate the effects of biokinetic, sorption and transport parameters on the FBR dynamics. More importantly, dimensional analysis and similitude techniques were employed to develop a process upscaling scenario. This process upscaling techniqueh was based on dimensionless groups associated with reaction kinetics, biofilm and particle diffusion transport as well as liquid film transport.

Laboratory-scale FBAR investigations were conducted to achieve maximum biological nitrification efficiencies. A series of batch studies were performed to determine the effects of several variables on the efficiency of the denitrification process including temperature, pH, total dissolved solids (TDS) and carbon-to-nitrogen (C:N) ratio. A series of chemostat tests were conducted to determine the Monod biokinetic parameters for the nitrification as well as the denitrification processes. It was found that nitrate was significantly more favorable as the main substrate for the denitrifying culture than nitrite, and that insufficient carbon source caused instability of the system due to inhibitory effect of nitrate and/or nitrite accumulation. The FBR denitrification experiments (first stage) were conducted under various conditions including different GAC quantities, hydraulic retention times and nitrate concentrations. These investigations showed that nitrate removal efficiencies as high as 96% to 100% could be achieved. Similar FBAR experiments conducted with sand showed that (i) completion of denitrification process took much longer time than with GAC with substantial nitrite accumulation, and (ii) significantly smaller biomass concentrations were obtained with sand as compared to GAC. It is important to note that the second FBAR achieved complete biological sulfate reduction to hydrogen sulfide, reducing the sulfate concentration almost to zero levels (third stage). The biofiltration process (fourth stage) effectively removed the hydrogen sulfide from the gas stream.

Model simulation studies demonstrated good agreement between the experimental data and model predictions. Model sensitivity analyses indicated that growth yield and maximum substrate utilization rate coefficients had the utmost influence on the process. The approach for upscaling the process from laboratory-scale to pilot-scale or full-scale is illustrated in this discussion.

Key words: biological denitrification, sulfate reduction, fluidized bed reactor modeling, process upscaling